DV Lottery 2012 July 15th statistical breakdown
- Thread starter Mele
- Start date
Thanks,
I don't hesitate to use the outcomes from your calculation at least for DV2012 AS region, since Bangladesh winners were less than around 6000. I think they have the right to set 5-7000 as preliminary cap on the basis of 7% rule, depending on demand/success rate of visa issuance in a particular country.
Here I show a part of my calculations for DV2012 Asia region.
fact1: totally 15002 winners were selected
fact2: 8562251/10102185=0.848, 84.8% of AS entries came from Bangladesh (B)
hypothesis from raevsky: 96.7% of Bangladesh applications were invalid for some reasons
question: Eating math for lunch is really fun, isn't it ?
1st round draw: Pick 15002 , 12722 (84.8 %) from B, 2280 from other countries (O), 420 (3.3%) of B is valid, total valid winner 2280+420=2700 out of 15002, CN reaches 15002
2nd round draw: Pick 15002 - (2280 + 420) = 12302, CN 15002 + 12302 = 27304, 344 from B , 1870 from O
3rd round draw: pick 12302 - (344+1870) = 10088, CN reaches 37392, etc etc
snip
By 26th round draw: total selectee 14916, 2319 from B, 12597 from other countries, CN reaches 82959,
Very good. Some AS winners having CNs higher then 70000 showed up in this forum.
It is close to actual B winners and this could also explain how to reach high CNs higher than winners in a single region.
Some of the reasons are listed in Edson's testimony.
They might have no valid photos, for instance - I am not sure phone books have ones. Or the same entries could be submitted twice by different companies (with pictures coinciding as files)
Just having my lunch now
Yes, except not 100% is valid from O as well. About 97.38% from Other countries are valid. Then after 26 iterations you would have 2358 from B, 12461 from O and 14819 total. And CN will be 84529.
It is very straightforward for AS because 3.3% is extremely low for B and 97.38% is extremely close to 100% for O. But the same thing could be done for other countries as well.
Last edited by a moderator:
BTW, with this type of selection algorithm we should see about uniform distribution of AS numbers throughout the whole interval from 1 to the maximum number. Do we see it uniform?
What I mean we should see about the same number of AS winners in the number intervals of the same lengths. So, the number of winners with numbers 1-10000 should be about the same as 10001-20000 and 20001 - 30000 and so on until the maximum possible number. Do we see that?
What I mean we should see about the same number of AS winners in the number intervals of the same lengths. So, the number of winners with numbers 1-10000 should be about the same as 10001-20000 and 20001 - 30000 and so on until the maximum possible number. Do we see that?
If you follow US Embassy Turkey DV interview schedule throughout a year, you would get an idea about this. As long as I followed that site for DV2010 by eye, I did not find any special/strange distribution, but of course, no guarantee. Unfortunately I didn't save it.
This suggests you need to put additional rule.
estimated invalid rates are too high (actual invalid cases are lower than your calculation) + another factor such as preliminary 7% cap.
What if invalid rates were set to lower than you calculated, and discontinue iteration of Ukraine and Uzbekistan when respective winners reached 5800 and 4800 (it will reduce total iterations and CNs more or less) , and continue other countries until winners became 30980 with the highest CN ranging 36000-40000?
Last edited by a moderator:
@Raevsky your is baseless for anyone to see,you just giving figures base on your own knowledge.Your calculation is not accurate cos failed to realise something.In Nigeria sometimes last november at the America embayy in Abuja,a senior US consulte manager in an open conference that in DV 2012 America government was going to selects over 6000 Nigerians in their lottery.Now we all are saying DOS is selecting winners on a fairness,no never have they selected winners on a fairness.It is a program stuff to select some particular winners cos there people out there even though if apply 100 times they will still not win.The machine is being program to select some particular winners.You talked about Nigeria being a fraudulent in DV submitting double entries,Man thats bullshit you are not a Nigerian you dont live in Nigeria to understand the system you think you can just sit down there in another country and give wrong details about Nigeria,guy you are absolutely wrong cos you have no idea or what so ever about Nigeria,dont pay much attention about the fraudulents in DV program about Nigeria,i know you can count Nigeria in other arears of fraudulent but in DV programs cos Nigerians are smart people to know the meaning of submitting double entries.
First of all great work people. One question here. Suppose you are selected by the lottery. Then, can I assume that Kentucky checks your entry if it is single or your photo is valid before sending you the NL1? If your entry is not valid they disqualify (they will not send you NL1) and somebody else takes your position?
Concerned_
Registered Users (C)
Correct, this has been discussed on many occasions.
Yes
Yes. But if it is a not exact duplicate, you will have an interview but the consul will see the other picture.
I do not know. Play with it yourself. It does not look like Ukraine or Uzbekistan is cut, 5800 and 4800 look to be their natural values without a cap. But I cannot just decrease the invalid rates - I calculate them.
Last edited by a moderator:
ok it makes sense but still something is not crystal clear to me. For DV 2012 they have selected 31,001 people from EU. Now, radovic claims that he saw somebody with EU36XXX. So this implies that at least 5,000 EU people have been disqualified from the lottery. Also it implies that statistically the same number of people should should exist (on average) between 26,000 - 31,000 and 31,000 - 36,000. Although I have seen many people with CN below 31,000, I haven't seen anybody above 31,000. Maybe I am missing something, but this case numbering system looks very complex...